Lentiviral Vif Degrades the APOBEC3Z3/APOBEC3H Protein of Its Mammalian Host and Is Capable of Cross-Species Activity

All lentiviruses except equine infectious anemia virus (EIAV) use the small accessory protein Vif to counteract the restriction activity of the relevant APOBEC3 (A3) proteins of their host species. Prior studies have suggested that the Vif-A3 interaction is species specific. Here, using the APOBEC3H (Z3)-type proteins from five distinct mammals, we report that this is generally not the case: some lentiviral Vif proteins are capable of triggering the degradation of both the A3Z3-type protein of their normal host species and those of several other mammals. For instance, SIV_mac Vif can mediate the degradation of the human, macaque, and cow A3Z3-type proteins but not of the sheep or cat A3Z3-type proteins. Maedi-visna virus (MVV) Vif is similarly promiscuous, degrading not only sheep A3Z3 but also the A3Z3-type proteins of humans, macaques, cows, and cats. In contrast to the neutralization capacity of these Vif proteins, human immunodeficiency virus (HIV), bovine immunodeficiency virus (BIV), and feline immunodeficiency virus (FIV) Vif appear specific to the A3Z3-type protein of their hosts. We conclude, first, that the Vif-A3Z3 interaction can be promiscuous and, second, despite this tendency, that each lentiviral Vif protein is optimized to degrade the A3Z3 protein of its mammalian host. Our results thereby suggest that the Vif-A3Z3 interaction is relevant to lentivirus biology.Lentiviruses are a unique class of complex retroviruses that encode a variety of accessory proteins in addition to the required Gag, Pol, and Env proteins. The archetypal lentivirus, human immunodeficiency virus type 1 (HIV-1), infects humans, but other members include simian immunodeficiency virus (SIV), bovine immunodeficiency virus (BIV), maedi-visna virus (MVV), caprine arthritis-encephalitis virus (CAEV), equine infectious anemia virus (EIAV), and feline immunodeficiency virus (FIV), which infect monkeys, cattle, sheep, goats, horses, and cats, respectively. The HIV-1 accessory protein viral infectivity factor (Vif) has been extensively studied because of its essential function in inhibiting the cellular antiretroviral human APOBEC3G (A3G) protein (43). HIV-1 Vif binds to human A3G (and other A3 proteins) and serves as an adaptor to link it to an ELOC-based E3 ubiquitin ligase complex (30, 51, 52). A3G is then polyubiquitinated and degraded by the cellular proteasome (7, 15, 29, 30, 43, 46, 52).Due to the potential therapeutic value of disrupting this host-pathogen interaction, a significant amount of work has been invested in defining the important contact residues between A3G and HIV-1 Vif. Primate A3G homologs have been useful tools in this effort, as many fail to be neutralized by HIV-1 Vif despite a relatively high degree of sequence similarity. For example, while HIV-1 Vif effectively neutralizes human A3G, it does not neutralize African green monkey A3G or rhesus macaque A3G despite 77% and 75% identity, respectively (4, 26, 27, 41, 51). The differential capacity of the HIV-1 and SIV_agm Vif proteins to degrade the A3G proteins of their hosts led to demonstrations that residue 128 is a key determinant: D128 made each A3G protein susceptible to HIV-1 Vif and K128 made each A3G protein susceptible to SIV_agm Vif (4, 26, 41, 51). This apparent on/off switch led to the prevailing model that the Vif-A3 interaction is species specific. However, even early data sets showed at least two hints that the story was more complex. First, the identity of the A3G residue 128 (K or D) does not diminish the interaction with the Vif proteins of SIV_mac or HIV-2 (41, 51). Second, SIV_mac Vif was shown to potently counteract the A3G proteins from rhesus macaque (as expected) but also those from human, African green monkey, and chimpanzee (27). Therefore, the implication from these studies is that the full nature of the A3-Vif interaction has yet to be elucidated.Although A3G has clearly served as the prototype for understanding the A3-Vif interaction, a growing number of studies indicate that other A3s are also capable of restricting lentivirus replication and interacting with Vif. A3G is one of seven human A3 proteins (A3A to -H) encoded in tandem on chromosome 22 (7, 16, 49). All but A3A have been implicated in the restriction of HIV-1 replication (reviewed in references 1, 10, and 45). For instance, human A3H has been shown to restrict HIV-1 replication and is susceptible to degradation by HIV-1 Vif (8, 37, 47). A3H is a Z3-type DNA deaminase characterized by a conserved threonine and a valine, in addition to the canonical H-x₁-E-x_23-28-C-x_2-4-C zinc-coordinating motif (23). The Z3-type deaminase is unique in that only one copy exists in all mammals whose genomes have been sequenced. It is encoded by a five-exon gene located at the distal end of each mammal''s A3 locus (adjacent to CBX7). Additional observations suggest that the Z3-type deaminases appear to have the capacity to restrict the Vif-deficient lentiviruses of their hosts. For example, African green monkey A3H can restrict the replication of SIV_agm and is susceptible to degradation by SIV_agm Vif, and the cat A3Z3 can restrict the replication of FIV and is susceptible to degradation by FIV Vif (33, 37, 48).Here, we take advantage of the fact that all sequenced mammals have a single A3Z3-type protein to test the hypothesis that these proteins are of general relevance to lentivirus restriction and to clarify the species-specific nature of the mammalian A3Z3/lentiviral Vif relationship. First, we ask if human, rhesus macaque, cow, sheep, and cat A3Z3-type proteins are all capable of retrovirus restriction. Second, we ask whether they are susceptible to Vif-mediated degradation in a host-specific manner. We show that each lentiviral Vif protein can indeed neutralize the Z3-type A3 protein of its host species. However, we were surprised to find that several of the Vif proteins, particularly SIV_mac and MVV Vif, can neutralize a broad number of A3Z3 proteins irrespective of the species of origin and overall degree of similarity. These data indicate that the A3-Vif interaction is more promiscuous than previously appreciated. Such broad functional flexibility may be relevant to understanding past retroviral zoonoses and predicting potential future events. We conclude that the A3Z3-Vif interaction is conserved on a macroscopic level, consistent with an important role in viral replication and particularly in species like artiodactyls and felines with fewer A3 proteins.